Save Job Back to Search Job Description Summary Similar JobsOwn evaluation science for frontier video & multimodal AI.Work directly with leading AI labs and publish impactful research.About Our ClientOur client is a globally recognised AI and data solutions organisation supporting the development of advanced generative AI and multimodal systems. The company partners with leading research labs and technology organisations to build robust data, evaluation, and benchmarking solutions for next-generation AI models.Job DescriptionDesign evaluation frameworks for video and multimodal AI models.Build benchmarks for temporal reasoning and video understanding.Create data specifications, annotation schemas, and quality standards.Conduct experiments linking data quality to model performance.Publish research findings, benchmarks, and methodologies.The Successful Applicant5+ years in video understanding or multimodal machine learning.Strong expertise in PyTorch and model evaluation methodologies.Experience with video-language and multimodal AI systems.Knowledge of Hugging Face, PEFT, and video data pipelines.Research publication record or notable open-source contributions.What's on OfferOpportunity to influence frontier AI research.Work with leading global AI research organisations.High-impact research ownership and visibility.Publication and thought-leadership opportunities.Exposure to cutting-edge multimodal and video AI challenges.Quote job refJN-102026-7125347Job summaryFunctionAnalyticsSub SectorMachine LearningWhat is your area of specialisation?Technology & TelecomsLocationNoidaJob TypePermanentJob ReferenceJN-102026-7125347Work from HomeWork from Home or Hybrid
